Towards on Develop a Framework for the Evaluation and Benchmarking of Skin Detectors Based on Artificial Intelligent Models Using Multi-Criteria Decision-Making Techniques

摘要: Evaluation and benchmarking of skin detectors are challenging tasks because multiple evaluation attributes conflicting criteria. Although several evaluating techniques have been proposed, these approaches many limitations. Fixing based on multi-attribute is particularly limited to reliable detection. Thus, this study aims develop a new framework for detection the basis artificial intelligent models using multi-criteria analysis. For purpose, two experiments conducted. The first experiment consists stages: (1) discussing development detector multi-agent learning different color spaces create dataset various space samples (2) testing developed according multi-evaluation criteria (i.e. reliability, time complexity, error rate within dataset) decision matrix. second applies decision-making (AHP/SAW, AHP/MEW, AHP/HAW, AHP/TOPSIS, AHP/WSM, AHP/WPM) benchmark results detector). Then, we discuss use mean, standard deviation, paired sample t-test measure correlations among ranking results.